argent-test-ui-flow: a skill that tests an app's UI flows with screenshots

argent-test-ui-flow makes your agent test an iOS or Android app by acting, checking the screen and reporting pass or fail with evidence.

Installs: 27.9K; on skills.sh
Stars: 3.1K; on GitHub
Updated: Oct 2026; last repo commit

The agent loops through five steps: take a baseline screenshot, find the target element, act, verify with the returned screenshot, repeat. It classifies each expected result as visual, structural, runtime or mixed, and picks evidence to match. For visual changes it uses screenshot-diff against the baseline.

It triggers when you ask to test a UI flow, verify login, test navigation or run QA steps. It needs the Argent tools and a booted simulator or emulator, which the per-platform setup skills handle. The final report states expected behavior, observed behavior, evidence and any blocker.

Good for

Developers who want a quick manual-style QA pass done by an agent
Teams checking login, navigation or visible UI changes
Developers verifying a UI edit looks right before and after

Install argent-test-ui-flow

Claude Code

Copy this prompt.

Install the argent-test-ui-flow skill in Claude Code. Run this command:

npx @swmansion/argent@latest init

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

This runs the Argent installer. It sets up the Argent tools (an MCP server) that these skills call, plus the Argent skills, in the editors you pick. Skills copied on their own with the skills CLI have no tools to call. The installer offers to add Argent tools to your editor's auto-approve list. The default Automatic skills option installs all Argent skills, not only this one. Telemetry is on by default and argent telemetry disable turns it off.

From the project README. Source.

What's in the skill

When it loads

Autonomously test an app UI (iOS or Android) by running interact-screenshot-verify loops using argent MCP tools. Use when testing UI flows, verifying login works, testing navigation, running UI test scenarios, manual QA steps, visible UI changes, or visual behavior.

Sections in SKILL.md

Platform-agnostic
1. Workflow
2. Template
3. Examples
Login flow
Scroll and navigation
Visual behavior check
Wait for a loading spinner
4. Recovery Pattern
Tips
Related Skills

Key facts

Author: Software Mansion (Argent)
Repo: software-mansion/argent, path packages/skills/skills/argent-test-ui-flow
Installs on skills.sh: 27,895 (as of Nov 2, 2026)
GitHub stars: 3,138 (shared repo, Nov 2, 2026)
License: Apache-2.0 for source code; some bundled binaries are proprietary
Last repo commit: Oct 31, 2026
Files: 1 file: SKILL.md (no bundled scripts)
Needs: The Argent tools (MCP server), set up by the installer in the box above

What it includes

screenshot at full scale, kept as a file
Uses scale 1.0 and includeImageInContext false so the large image is saved, not loaded into the chat.
Files · 1. Baseline screenshot
describe, debugger-component-tree or screenshot
React Native apps use debugger-component-tree first. Other screens use describe. Screenshot is the fallback.
Read · 2. Find target
gesture-tap, gesture-swipe, keyboard, button
Each call returns a screenshot automatically.
3. Interact
Visual, structural, runtime and mixed checks
screenshot-diff for looks, describe for structure, view-network-logs and debugger-log-registry for runtime.
Read · 4. Verify
Login flow, scroll and navigation, visual check, wait for a spinner
The login example uses a {{secret:APP_PASSWORD}} placeholder so no plaintext password is typed.
Templates and examples
Re-discover after a missed tap, stop after two failures
Uses await-ui-element instead of blind delays. Saved e2e tests follow argent-create-e2e for repair.
Recovery pattern
Seven companion skills
argent-device-interact, argent-screenshot-diff, the two setup skills, argent-react-native-app-workflow, argent-metro-debugger, argent-create-e2e.
Related skills

From the official docs

Permissions

Medium risk

The agent taps and types in your app, so it can submit forms or change data in the app. Use a test account on a simulator and keep passwords in a secrets file.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

The agent acts inside your app
It taps, types and navigates, which can submit forms or change data. Use a test account and a throwaway simulator. github.com
Installer can auto-approve Argent tools
`argent init` asks whether to add Argent tools to your editor's auto-approve list and its prompt defaults to yes. Say no if you want a confirmation on every tool call. github.com
Passwords come from local env or secrets files
SKILL.md says to use a {{secret:NAME}} placeholder. The value comes from an ARGENT_SECRET_ variable or a secrets file in the project or your home folder. Put only test credentials there. github.com
Argent telemetry is on by default
The README says telemetry is opt-out. Run `argent telemetry disable` to turn it off; the privacy notice says it collects no source code, file paths or tool inputs, and records a hashed installation identifier derived from a device identifier, never the raw one. github.com

Technical details

Talks to	Argent telemetry (on by default)
Sign-in and keys	The skill text does not use an account or API key.

Questions

What does argent-test-ui-flow do?

It makes your agent test an app's UI by following a flow on a simulator or emulator. It acts, checks a screenshot or the element tree, and reports pass or fail with the evidence it used.

How do I install it in Claude Code?

Use the install box above. It runs the Argent installer, which sets up the device tools this skill calls along with the skills.

Does it work on a real iPhone?

SKILL.md has a note for a physical iPhone: read argent-ios-device-interact first, call launch-app before anything, and note that describe fails while the app is backgrounded.

Is it safe to run?

The agent taps and types in your app, so use a test account on a simulator. It bundles no scripts. The installer can also add Argent tools to your editor's auto-approve list, which you can decline.

How does it type a password?

It uses a {{secret:NAME}} placeholder in keyboard so the value is filled in by the tool server and never enters the chat. The value comes from an ARGENT_SECRET_ variable or a secrets file.

How is it different from argent-device-interact?

argent-device-interact explains how each tool works. This skill is the test loop built on top of it: baseline, act, verify and report.

Can I replay a flow later?

Yes. SKILL.md says to record a flow that will be repeated with the argent-create-e2e skill as a .yaml script, then replay it with a single e2e-execute call.

Information

updated Nov 2, 2026

Repository: software-mansion/argent
Folder: packages/skills/skills/argent-test-ui-flow

software-mansion/argent skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 5

SKILL.md · github.com
Argent README · github.com
Argent installer: editor adapters · github.com
Argent installer: auto-approve step · github.com
init-skills.ts · github.com